Myth of the Revolving Door Superintendency

Contrary to perception, tenure runs much longer than most believe, a research review finds by Gerald Natkin, Bruce Cooper, Lance Fusarelli, James Alborano, Arthur Padilla and Sujit Ghosh

For a decade or more, conventional wisdom has held that the superintendency, particularly in urban districts, is an unstable profession in which incumbents typically come and go every few years. This perceived phenomenon has been dubbed the “revolving-door superintendency.”


But does the revolving-door superintendency really exist?

Schools and school systems today are under extraordinary pressure to address a variety of issues: implementing standards and high-stakes testing; shifting instructional/pedagogical strategies; eliminating the minority/majority achievement gap; using technology productively; attending to diversity concerns; enhancing professional development; and changing teacher certification requirements.

Successful change on so many fronts requires constancy of purpose and stable and predictable leadership over a sustained period of time. Michael G. Fullan and Suzanne Stiegelbauer, writing a decade ago in The New Meaning of Educational Change, estimated that successful reforms require five years or more of a superintendent’s attention. Thus, if tenures of two or three years are as widespread as commonly believed, comprehensive school reform might be an elusive goal.

A Critical Issue?

First-hand experience shows that lack of stability at the top can indeed be an obstacle to successful innovation.

One of us was a central-office administrator in a large, metropolitan district that experienced repeated leadership changes during a six-year period. There was often a sense of instability, that decisions and mandates were only temporarily valid.

For example, he led a team in planning the district’s early efforts in instructional technology. After two years of work, the team presented the plans to the school board, which approved a substantial budget for pilot efforts. Within a few months, however, a new superintendent, for whom classroom technology was not an immediate priority, was hired. Consequently, the pilot plans were shelved for almost two more years.

Policies and previous decisions were not the only things to change under each new leader. The district also experienced frequent turnover of administrative and managerial staff. This is perhaps the most unsettling kind of instability in organizational life because factionalism and unhealthy competition can emerge, morale may suffer, and survival can easily become more important than productivity or school improvement.

While frequent superintendent turnover can be a serious problem, the amount of attention the issue receives in both popular and academic publications gives the impression that the revolving-door superintendency is so widespread that it may be approaching critical proportions.

For example, in their August 1994 article in Executive Educator, “Why the Superintendency Is Becoming a Revolving Door,” educational leadership professors Jack McKay and Marilyn Grady claim that the “three-year cycle of dismissal, search and selection, reorganization and dismissal again [was] the greatest single hindrance to improving the quality of our schools.” Yet they provided no factual evidence to indicate this revolving door cycle is the norm. Is the situation really that bad?

Of course, some superintendents do leave their positions after a short time on the job. Everyone can cite an example of a superintendent arriving in a district heralded as its savior, enjoying strong support from the school board and the local power structure to increase standards, improve programs and raise sagging test scores, only to run afoul of a teachers’ union or board member with an ax to grind. Within two or three years, the superintendent has been fired.

Such well-publicized stories of involuntary turnover grab our attention and perhaps reinforce the belief that brief tenures are the norm. But just how typical are such cases?

Research on Turnover

The prevalence of frequent superintendent turnover can’t be established by newspaper articles or even academic case studies. What’s required is well-designed quantitative research based on adequate sampling. To put the issue on a more factual basis, we reviewed available quantitative studies and undertook our own extensive survey of school district archival records.

Before we review the quantitative research findings, we first need to address a problem that can arise in interpreting studies of turnover. In this type of research, we are usually concerned with average total tenure, the amount of time superintendents remain in a single position. Total tenure data can only be obtained for superintendents who have completed their final contracts in a given district; it is not available for incumbents, who may remain in office for several more years.

Some studies, including two we discuss here, focus heavily or exclusively on incumbents. There is nothing inherently wrong with such studies, but they need to be interpreted very cautiously because their tenure statistics may seriously underestimate average total tenure and appear, incorrectly, to justify belief in the revolving-door superintendency.

In a 1996 study titled “When Is Tenure Long Enough?” Gary Yee and Larry Cuban determined that the mean tenure for superintendents of the nation’s 25 largest districts who were in office in 1990 was 5.76 years—more than twice as long as the 2.5 years usually claimed. They also found that tenure has declined significantly since the middle of the last century when it averaged 13-14 years and that although tenure in 1990 was at its lowest recorded point, there have been previous cycles of increase and decrease, even in recent decades.

Eighty-eight percent of the superintendent tenure records Yee and Cuban (himself a former superintendent), reviewed for 1990 were complete, so their results give a quite accurate picture of tenure in the nation’s largest districts at that time.

More recently, Bruce Cooper, Lance Fusarelli and Vincent Carella of Fordham University conducted a large-scale, AASA-supported survey of incumbent superintendents’ opinions, skills, career concerns and future interests.

Although turnover was not the main focus of the 2000 study, titled “Career Crisis in the Superintendency?” participants were asked how long they had been in their current positions. Overall, the 1,688 respondents had held their positions for an average of 7.25 years, while superintendents of the 88 largest districts at that time had served an average of 4.71 years. Since the data for this study are from incumbents, the statistics no doubt underestimate total tenure.

The Study of the American School Superintendency 2000, an AASA-sponsored 10-year study of the profession directed by Tom Glass, a professor of educational leadership at University of Memphis, estimated that the tenure for the 2,232 superintendents for whom Glass and his colleagues, Lars Björk and C. Cryss Brunner, had the necessary data averaged between five and six years.

Finally, a survey conducted in June 2000 by the National School Boards Association’s Council of Urban Boards of Education further confirms these findings. Among 77 CUBE member school districts that responded to the survey, the most recent superintendent to complete his or her superintendency served an average of five years.

The only data we found that supports the notion of the revolving-door superintendency came from a study conducted by the Council of Great City Schools. As part of a larger survey, member districts were asked how long their current superintendents had been in their positions. Mean tenure for the 48 districts that responded was 2.33 years. These incumbent-based results necessarily underestimate total tenure. In addition, because the council’s member districts are, on average, many times larger than the nation’s other urban systems, it would be risky to generalize the findings.

A Retrospective Study

Although the quantitative evidence in these published studies already seemed to refute the concept of the revolving-door superintendency, we decided to conduct a new, large-scale study to provide accurate statistics on the magnitude of superintendent turnover and show how it may have changed.

With assistance from AASA, we undertook a 25-year retrospective study of tenure in 468 school districts. National samples were selected at random from districts of various student population sizes ranging from fewer than 1,000 to more than 50,000. In part to crosscheck national results, we also sent surveys to all school districts in North Carolina.

The survey requested information about each superintendent holding office in the district since 1975, including the year the superintendent was first hired as well as his or her tenure. We also requested data about the size of the district (number of students and number of schools) and its demographic type (rural, small town, suburban, urban).

The survey response rate was 44 percent in the national sample and 80 percent in the North Carolina study. Data were obtained for 1,103 superintendents; 274 of them were still in office at the time the study was conducted.

Contrary to earlier studies, we report median rather than mean tenure because means (or averages) can be distorted by the few extreme tenures of 20 or 25 years still sometimes found among superintendents. The data were analyzed by Cox proportional hazards regression, a statistical technique that allowed us to include data for both incumbent and former superintendents and weight each appropriately.

A Factual Look

Our large-scale study provided a very detailed long-term look at superintendent tenure. Among the findings were these:

Median tenure between 1990 and 1994 was 6.5 years—more than twice as long as we are typically led to believe.

Definitive data for the most recent years are unavailable because many of the superintendents hired since 1995 are still in office. However, the data we do have for this period do not suggest a downturn in tenure.

Turnover has not increased markedly since 1975.

Median tenure was 7.5 years in 1975-1979 and 6.5 years in 1990-1994. Thus, median tenure declined by only about one year over the period we studied—statistically a nonsignificant change.

Turnover isn’t related to the size of the district’s student enrollment.

Perhaps the largest districts in the nation, such as many of those included in the Council of Great City Schools study, have significantly higher turnover than others. However, for the district sizes we studied, from fewer than 1,000 students to more than 50,000 students, there wasn’t a significant difference.

Turnover isn’t related to the demographic setting of districts.

Rural, small town, suburban and urban settings all showed approximately the same turnover patterns.

Thus, with the exception of the CGCS survey, the statistical results of these studies are consistent: Public school superintendency is not a revolving-door profession, whether in small districts or large, urban or rural, in the early 1970s or the 1990s, nationwide or in a single Southeastern state.

Consequences of the Myth

Glass, Björk and Brunner, the co-authors of AASA’s most recent 10-year report, assert that the widespread belief in the revolving-door superintendency has damaged the image of the superintendency. That, we argue, is only one of its many negative consequences.

To better understand the implications of our research findings, we shared them with selected superintendents and school administrators, as well as with colleagues in regional education organizations, universities, state departments of education and professional organizations, all of whom work with superintendents.

These professionals offered a wealth of suggestions about how the myth of the revolving-door superintendency has affected education. Many of their insights are reflected here.

The myth influences the actions of educators, superintendents, board members, policymakers and other important educational stakeholders. Indeed, belief that turnover is excessively high affects every aspect of a superintendent’s career, from initial training to reception on the job.

Superintendents, believing they will likely be in a position for a short time, may be reluctant to undertake major reforms, and instead focus their efforts on objectives that can be accomplished within two or three years. This attitude is encouraged even in professional literature, in which superintendents are sometimes advised to keep their bags packed, rent rather than buy new homes, quit their jobs rather than fight, and expect and prepare for failure.

School boards, administrative staff, teachers and the community power structure also may be conditioned by the revolving-door myth. In particular, the "this too shall pass" attitude with regard to improvement initiatives may be reinforced by the expectation that a new superintendent, with a different set of priorities and mandates, soon will be in place.

Finally, as Glass and his colleagues suggest, the myth of high turnover may detract from the image of the superintendency as a profession and lead the public and editorial writers to view school system CEOs as "hired guns." This attitude, in turn, may reflect negatively on the entire enterprise of public education.

A Different Perspective

We have documented that superintendents typically remain in their positions more than six years—certainly enough time for them to oversee the development of long-term plans and the implementation of major school improvement initiatives.

However, the education community and the news media need to focus on undoing the damage caused by the longstanding, erroneous belief in the instability of top leadership. This effort is vital to making the superintendency more attractive, to assuring an adequate supply of top system leaders for the future, and to communicating to educators and the public that superintendents can provide a sustained force for educational change.

Jerry Natkin is a research scientist with SERVE, the regional educational laboratory serving the Southeast, at the University of North Carolina at Greensboro, P.O. Box 5367, Greensboro, NC 27403. E-mail: jnatkin@serve.org. Bruce Cooper is a professor and vice chair of administration, policy and urban education at Fordham University Graduate School of Education. Lance Fusarelli is an assistant professor of administration, policy and urban education at Fordham. James Alborano is an adjunct professor of professional and secondary education at East Stroudsburg State University. Art Padilla is a professor of business management and Sujit Ghosh is an associate professor of statistics, both at North Carolina State University.